• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

Xbox "Advanced Tech Engineers" discuss tech specs

tetsuoxb said:
No it isnt. What do you expect them to say - Sony kicked our ass? THEY WORK FOR MS.

Some of the stuff they say makes a lot of sense. I'm not taking this wholesale, but there seems to be a lot of truth to what they say.
 
I'm listening now.

It's obviously biased but it's also not too outlandish.

We'll know in 18 months which system is more powerful anyway.
 
Fatghost28 said:
I'm listening now.

It's obviously biased but it's also not too outlandish.

We'll know in 18 months which system is more powerful anyway.

Yeah, it's obvious they have a bias (they worked on the thing), but some stuff about the X360 cores vs. PS3's cell makes a ton of sense.
 
Pedigree Chum said:
Yeah, it's obvious they have a bias (they worked on the thing), but some stuff about the X360 cores vs. PS3's cell makes a ton of sense.


Except that more "general" CPU power doesn't necessarily matter anyway. Saturn had more general CPU power than PS1 but the PS1 had seriously outclassed the Saturn where it counts.

PS3 may be less useful as a general computer but superior as a games machine. We won't really know which design philosophy is superior until we've had a round of software on both machines to compare against.

That said, the deck certainly seems stacked in favor of PS3 anyway, just with the media choice.
 
tetsuoxb said:
No it isnt. What do you expect them to say - Sony kicked our ass? THEY WORK FOR MS.


Well the fact that it's public information on the net that a thousand fanboys and a few dozen intelligent people that actually know what they're talking about will be sifting through their comments means that can't make anything TOO outlandish. I'm extremely curious as to what Vince and arha have to say about these things.

Their main points seem to be the edram on the X360 and the simplicity of the SPEs on the PS3 side. The hilarious thing is that if the RSX ends up including some edram then that'll be a huge boost for the PS3 (also depending on how "smart" the edram is). As for the lack of CACHE on the SPEs, that's rather surprising. I knew they had local memory, but I thought that included cache as well, or at least some kind of scratchpad like in the PS2.
 
Interesting, thanks.


the techs go to work tearing down CELL and also the bandwidth limitations of RSX.


it's biased yes, but what else would you expect from guys that actually worked on Xbox 360 hardware? :)
 
Ever read Dilbert, when the marketing clowns try to get the engineers to help? I love those.

Anyways .. someone push the fast forward button on life ... I'm sick of waiting.
 
"We did the most accurate comparison"

Yeah, whatever. That guy has the most pseudo-intellectual gay voice with those drawn-out words I've ever heard. The criticism of CELL and the SPEs is such horsehit. Fuck you Microsoft, and fuck J.Allard too.
 
CosmicGroinPull said:
So they're getting engineers to do marketing now? :lol Very innovative. I approve.

Major's blog is actually pretty cool. He's the Director of XBox Live programming, pretty high up in Microsoft, and he's really interesting in communication technology as a way to bring groups together. His weekly blogcast is usually very interesting, and he even takes calls on Skype on occasion and adds them in.

He asked people to post questions about those comparisons between the PS3 and the 360 in his blog and got a few hundred, and he brought in the authors of the PR in to answer some of those quesitons. Pretty cool if you ask me - sure it's PR but it's PR straight from the engineers who worked on the system - makes it a wee bit more credible to my ears.
 
Vince said:
"We did the most accurate comparison"

Yeah, whatever. That guy has the most pseudo-intellectual gay voice with those drawn-out words I've ever heard. The criticism of CELL and the SPEs is such horsehit. Fuck you Microsoft, and fuck J.Allard too.

oh come on like Sony isn't doing it as bad or even worse.

"xbox 1.5" :\

it's obvious where your loyalties are... how could anyone trust you for accurate hardware information?
 
Vince said:
Why, so he can try slitting his wrists with the nearest plastic spoon after listening? And yes, it's that fucking bad.


NO so he can try to balance the bullshit that I'm sure has been spread liberally.
 
Razoric said:
oh come on like Sony isn't doing it as bad or even worse.

"xbox 1.5" :\

Write, since a comment of "XBox1.5" is equivalent to this form of marketing that does nothing but take advantage of people's supposed inability to understand what these two guys are saying and just agree.

Let me fill you in, this is more biased than the most biased thing I've likely posted. And I know how bad I've been, but this is just that bad.
 
Vince said:
"We did the most accurate comparison"

Yeah, whatever. That guy has the most pseudo-intellectual gay voice with those drawn-out words I've ever heard. The criticism of CELL and the SPEs is such horsehit. Fuck you Microsoft, and fuck J.Allard too.

I love comments like these. I read them and hear Luke shrieking, "That's not true! That's impossible!!!!"

Seriously - if what they say is such bullshit, then tell us HOW it's bullshit. They made some very, very specific claims and I'm interested in a coherent counter point. Just calling it crap does nothing. What they said sure sounds a lot less like BS that what Sony has cranked out.
 
Vince said:
"We did the most accurate comparison"

Yeah, whatever. That guy has the most pseudo-intellectual gay voice with those drawn-out words I've ever heard. The criticism of CELL and the SPEs is such horsehit. Fuck you Microsoft, and fuck J.Allard too.
Xbox ATG has profiled and optimized hundreds of current and next-gen Xbox/360 games in development over the past few years.

How many current and next-gen games have you worked on, Vince?
 
GhaleonEB said:
Seriously - if what they say is such bullshit, then tell us HOW it's bullshit. They made some very, very specific claims and I'm interested in a coherent counter point. Just calling it crap does nothing. What they said sure sounds a lot less like BS that what Sony has cranked out.


Which is why I want my Pana!
 
Vince said:
"We did the most accurate comparison"

Yeah, whatever. That guy has the most pseudo-intellectual gay voice with those drawn-out words I've ever heard. The criticism of CELL and the SPEs is such horsehit. Fuck you Microsoft, and fuck J.Allard too.

With a comment like this, I'd guess you would speak from experience about Cell and programming it for games, right? Why is what they've said horseshit? I wanna know. There's gotta be some perspective on their comments, as it's unlikely that Sony will bother putting up similar early analysis of their competitor in X360. Basically, would PS3 be better for general purpose programming that would make making games on it easier? I'm just boiling down what I've gotten from that blogcast. Their statements make it clear that they feel that X360 is easier to extract high performance from than PS3, due to a more balanced design aimed at the target application of the hardware. They feel that PS3 is focused on higher performance potential at the cost of ease of use.

I'm only asking you because you seem to be a PS3 developer, based on your comment.
 
GhaleonEB said:
Major's blog is actually pretty cool. He's the Director of XBox Live programming, pretty high up in Microsoft, and he's really interesting in communication technology as a way to bring groups together. His weekly blogcast is usually very interesting, and he even takes calls on Skype on occasion and adds them in.

He asked people to post questions about those comparisons between the PS3 and the 360 in his blog and got a few hundred, and he brought in the authors of the PR in to answer some of those quesitons. Pretty cool if you ask me - sure it's PR but it's PR straight from the engineers who worked on the system - makes it a wee bit more credible to my ears.



nice post Ghaleon.

but they did not get technical enough for a truly deep understanding of either machine, plus they said that Sony's PS2 demos from years ago were prerendered when in fact they were realtime. and various other things that were either wrong, or slanted.

of course though, nobody is going to take my word over some Microsoft engineers :|

overall, it was a mildly interesting audio, its always fun to hear with your own ears (rather than reading text of articles or message board posts) what we've been discussing for a long time now.


I am waiting to hear DaveBaumann's (of Beyond3D) conference call with the ATI architects of the Xenos GPU. that should be REALLY interesting.
 
rastex said:
Well the fact that it's public information on the net that a thousand fanboys and a few dozen intelligent people that actually know what they're talking about will be sifting through their comments means that can't make anything TOO outlandish.
Meh, why would they need to? Surely you've seen Shopping Channel advertisements before, this was exactly the same thing. All they needed was avoid everything that may not paint them in best light, and embelish things they DO talk about.

The hilarious thing is that if the RSX ends up including some edram then that'll be a huge boost for the PS3 (also depending on how "smart" the edram is).
Well until we know more about RSX it's hard to say what it would need. What if it's a TBDR? :P Or has something obscene like 1MB of framebuffer cache... or... etc.

As for the lack of CACHE on the SPEs, that's rather surprising. I knew they had local memory, but I thought that included cache as well, or at least some kind of scratchpad like in the PS2.
And this just goes to show how these marketting stunts ultimately serve no other purpose then confuse people (as opposed to educating them).
Scratchpad IS localmemory (or if you want to get technical - SRAM minus the caching logic).
 
How about: The two consoles are both approx. as powerful as each other. There's no point in arguing over this crap. As soon as they get released people will be focusing on what really counts: the games and not stupid tech numbers. It's like those ghey "ilovebees" and "ourcolony.net" things. People were going crazy over them for no real reason since once Halo 2 was released/Xbox 360 was revealed there was no need to wait time on stupid advertisements hidden as games.
 
midnightguy said:
nice post Ghaleon.

but they did not get technical enough for a truly deep understanding of either machine, plus they said that Sony's PS2 demos from years ago were prerendered when in fact they were realtime. and various other things that were either wrong, or slanted.

of course though, nobody is going to take my word over some Microsoft engineers :|

overall, it was a mildly interesting audio, its always fun to hear with your own ears (rather than reading text of articles or message board posts) what we've been discussing for a long time now.


I am waiting to hear DaveBaumann's (of Beyond3D) conference call with the ATI architects of the Xenos GPU. that should be REALLY interesting.


just out of curiosity... are you megadrive on B3D?
 
Fafalada said:
And this just goes to show how these marketting stunts ultimately serve no other purpose then confuse people (as opposed to educating them).
Scratchpad IS localmemory (or if you want to get technical - SRAM minus the caching logic).


I agree, which is why I posted. This is blatently abusing the ignorance of people in the same way that Intelligent Design followers and Creationists do in the Biological community.

On the topic of CELL and the SPEs, just read the papers yourself:

NO so he can try to balance the bullshit that I'm sure has been spread liberally.

They blatently lie on the topic of CELL and it's processing potential, why the architecture is designed as it is and why they did it.

If you want a genuine research report on CELL, check out Power Efficient Processor Design and the Cell Processor for the philosophy behind the CELL design, and the SPEs.

Or P. Hofstee's paper on CELL which is more technical.

M. Gschwind on Exploring Real Time Multimedia Content Creation in Video Games
 
I have to say even though I despise this viral marketing bullshit I absolutely love the effort out of MS here. I never imagined it getting quite this carny. "Major Nelson" is like PT Barnum... with an engineering degree and a D&D die in his back pocket.
 
While I'm sure there are things they are saying which are total BS and will be proven as such in a year. I find it just as funny that some people just believe everything the other company is saying as 100% fact.

To me both these consoles look powerful. There's some advantages and disadvantes to both in comparison to the other. Why others can't accept this based on what we know now just proves how irrational they are.
 
Interesting interview. They kind of blew off the advantages of Cell.

And with an Nvidia GPU why wouldn't PS3 have enough power to render games in 1080p?
 
GhaleonEB said:
I love comments like these. I read them and hear Luke shrieking, "That's not true! That's impossible!!!!"

Seriously - if what they say is such bullshit, then tell us HOW it's bullshit. They made some very, very specific claims and I'm interested in a coherent counter point. Just calling it crap does nothing. What they said sure sounds a lot less like BS that what Sony has cranked out.
And I love comments like these that give more credence to the comments of a few MS engineers that haven't touched finalized PS3 hardware or seen full specs but have no problems calling Sony's statements BS outright while accusing another poster of not fulfilling their burden of proof. You have to be badly steeped in corporate fanboyism not to realize how many double standards you just waved around.
 
kaching said:
And I love comments like these that give more credence to the comments of a few MS engineers that haven't touched finalized PS3 hardware or seen full specs but have no problems calling Sony's statements BS outright while accusing another poster of not fulfilling their burden of proof. You have to be badly steeped in corporate fanboyism not to realize how many double standards you just waved around.

Hey kaching... WARHAWK!!!!
 
The major problem with this is that their clearly making educated guesses re: CELL, SPEs and RSX. They reach conclusions and back them up with half truths wrapped in a veil of jargon. They don't explore any advantages of the PS3s design, or weaknesses of X360s.

To me both these consoles look powerful. There's some advantages and disadvantes to both in comparison to the other. Why others can't accept this based on what we know now just proves how irrational they are.
Well, its fun to draw lines in the sand and argue. Also, with MS and their fans on the backfoot, it makes it more interesting. Just as hard as it is to accept that they are both powerful, how hard would it be for you to accept that PS3 might have a (discernable) advantage? Well, they admit as much in the blogcast, they just discard it.
 
Dude, I KNOW. You Rock...but what about that action game for the PSP you kept showing that one FMV still for? ;)
 
kaching said:
And I love comments like these that give more credence to the comments of a few MS engineers that haven't touched finalized PS3 hardware or seen full specs but have no problems calling Sony's statements BS outright while accusing another poster of not fulfilling their burden of proof. You have to be badly steeped in corporate fanboyism not to realize how many double standards you just waved around.

Actually, I don't really buy the story either side is telling; I'm not nearly enough of the techie to decipher it myself and I don't presume to be.

I call Sony's stuff BS because their entire approach has caused them to lose credibiltiy in my eyes. They put out only a handful of off the wall specs, then show a ton of rendered footage deliberately to give the impression that it's realtime. 90% of what they SHOWED was bullshit, so I sure as hell don't trust what they are telling me.

Meanwhile, MS comes to E3 and says, "our system is really powerful, but we've only got Alpha hardware out there so far. Here's what we've got." It's an entirely more HONEST approach.

That said, as I've said over and over and over and over, I'll be really suprised if the PS3 isn't the more powerful machine. If it's not, not only did MS pull off a miaracle but lots of (incompetant) heads are gonna roll over at Sony.

I'm reserving judgement on either system's power until folks like Pana and sites lie Ars Technia put out their analysis.
 
Striek said:
The major problem with this is that their clearly making educated guesses re: CELL, SPEs and RSX.
They admitted at the beginning of the interview that that was the case. Sony did the same thing despite of the fact they don't have final 360 hardware.
 
MightyHedgehog said:
Why is what they've said horseshit? I wanna know.

Criticism of the CELL architecture's SPEs:

1) They don't do proper FP rounding

He's refering to the fact that the SPEs aren't fully IEEE754 compliant since they round to zero, just like the Emotion Engine does. This is is because for real-time applications (eg. games) the equilibrium point between [die] area and featureset that yeilds the highest preformance is what they chose.

I find this ironic considering ATI's current support of a [proprietary] 10-10-10-2 [HDR] format in the X360 GPU. Why would they chose some non-standard FP10 format when nVidia uses FP16 and FP32?

Oh, right... because this is nothing but marketing bullshit. In both cases, ATIs FP10 and STI forgoing IEEE754 blending, they chose the correct and logical design.

2) They don't have a branch predictor

Again, this is to maximize the preformance by finding the best point of equilibrium in design. it's a trade-off between more hardwired functionality that limits preformance and limits a developer's options or the choice of software, which entails added flexibility, added preformance but at the cost of doing the work in the tools and coding.

P. Hofstee said:
Whereas modern hardware branch prediction structures do not do well on performance per transistor, software branch prediction tends to be an inexpensive means to gain a substantial performance improvement. In order to support high-frequency deeply pipelined designs efficiently, a branch hint instruction, present in the code well before the actual branch instruction allows for “just in time” prefetching of the instructions at the branch target. In combination with loop unrolling and interleaving, compilers can achieve near optimal performance even on deeply pipelined high-frequency processors.

3) Their DP math is extremely slow

The SPEs have a IBM designed DP unit that is fully fully IEEE854 compliant. This is something that the Emotion Engine got away without completely and is nearly irrelevent for Real-Time applications. Now, if you demand that your games simulate alot of non-linear multivariable dynamics or statistical mechanics problems, then that's one thing. The question to be asked back is, Do you propose that all smoke and water effects on PlayStation3 must take into account fluid dynamics and model turbulant flow physically accuratly? Liberal use of the Navier-Stokes equations perhaps? it just doesn't make sense.

While on the topic of this "extremely slow" DP operation in the CELL architecture, let it be noted that a single PS3-CELL which will retail for $300 is 3 (THREE) times faster at DP Floats than the NEC SX-6 that is a custom IC designed for use in the Earth Simulator -- the 30TFlop super computer.

4) The biggest problem is that instead of a cache they have a Local Memory

Uh, no shit sherlock. This is basically necessitated by the above SPE choices we've talked about and the desire to have General-Computation on the SPEs. Besides, this cache (especially in what a console dev enviroment affords) is extremenly ineffecient proposition:

P. Hofstee said:
Almost all modern high-performance microprocessors devote the majority of transistors to caches and other structures in support of maintaining the illusion of a large, uniformly accessible, memory. As a rule of thumb cache miss rates are inversely proportional to the square root of their size, thus to first order confirming GelsingerÂ’s law. The illusion of a flat and uniformly accessible memory is however increasingly costly to maintain. Latencies to main memory, in spite of designerÂ’s best efforts range in the several hundreds of cycles and approach a thousand cycles in multi-GHz SMP systems. With such a large penalty associated with a cache miss, managing memory locality becomes the main factor determining software performance, and writers of compilers and highperformance software alike spend much of their time reverse-engineering and defeating the sophisticated mechanisms that automatically bring data on to and off the chip. Given the large number of transistors devoted to these mechanisms this is an unsatisfactory situation.
 
kaching said:
Dude, I KNOW. You Rock...but what about that action game for the PSP you kept showing that one FMV still for? ;)

Patience my friend. I know the E3 PSP showing was a huge letdown but brighter days are just around the corner... ;-)

My two cents on the PS3 vs. X360 architecture debate:

MS is deeply worried that they have lost a major advantage amongst Western developers - namely that the CELL/RSX tandem combined with OpenGL will attract traditional PC developers who avoided PS2 development altogether this gen. The very fact that UE3 is up and running on alpha PS3 kits and Tim Sweeney is openly embracing CELL/RSX is a dagger to the heart of MS's ambitions to make X360 easier to developer for via XNA and thus keep the major PC development teams firmly on their side. The fact that Sony will lock up Japan's biggest teams and exclusives and possibly capture studios like Bioware, Epic and Id has to be greatly worrying to MS. Hence, hitherto unseen damage control from Redmond HQ.
 
dorio said:
They admitted at the beginning of the interview that that was the case. Sony did the same thing despite of the fact they don't have final 360 hardware.
Well, Sony really only compared numbers AFAIK. So, its not really the same thing.
 
MassiveAttack said:
Patience my friend. I know the E3 PSP showing was a huge letdown but brighter days are just around the corner... ;-)

I hope you're right. DS totally blew me away at E3, and PSP...well...let's just say it was one of my E3 lowlights.

MassiveAttack said:
The fact that Sony will possibly capture studios like Bioware, Epic and Id has to be greatly worrying to MS. Hence, hitherto unseen damage control from Redmond HQ.

Is there any specific news you're privy to that leads you to believe that, for instance, BioWare, Epic or Id would be supporting PS3 more than Xbox360? Or do you just mean they'll be supporting PS3 more than they supported PS2 last gen?
 
Well I just started reading the white papers and I already have to agree with Vince.

It seems like the main advantage MS guys are stressing is that on the 360 you can write lazy code and still get halfway decent performanc/ (I guess this is their "general purpose CPU power").

Cell has a simplified CPU design and you code it "at the metal" resulting in 2x performance advantage.
 
I will sum it up - Face it people The PS3 will be a bit more powerful than the X360, probably about 20 %.

Those who think the X360 will be more powerful than PS3 are living a fanboy delusion.
Those who think the PS3 is much more powerful than X360 are living a fanboy delusion.

I think most of the anquish seen in these tech debates is from the PS3 fans who bought into the '2x better' and dont want to come back down to reality.

And MS fans? Come one guys the X360 will never be more powerful than the PS3 even if the Cell is a bit less efficient for game programming.
 
seismologist said:
It seems like the main advantage MS guys are stressing is that on the 360 you can write lazy code and still get halfway decent performanc/ (I guess this is their "general purpose CPU power").

And it's actually more hypocritical than that. They attempt to take the developer's high ground so-to-speak, yet when they had to choose between helping developers by sacrificing preformance for ease of programming, they chose preformance.

The X360's CPU doesn't support OoOE for the same reasons and the same mentality that guided CELL development. This is nothing but pathetic, and that guys voice is still f-ing gay.


aaaaa0, why does it matter? We both know the answer. Does it change what I've posted or the positions I've taken? Does my working in the medical field and doing research with specialization in biochemistry and neurology change my ability to comprehend and articulate this? I think not, instead of attacking who I am of which you don't know (my nickname is Vince BTW) attack what I said... or atleast how I said it. :)
 
Striek said:
Well, Sony really only compared numbers AFAIK. So, its not really the same thing.
So did these guys. They mentioned the numbers that sony produced even given them the benefit of the doubt on a couple numbers.

In the end, we'll won't truly know how any of this plays out in the realworld until a year after the latest console is released. There is a steep learning curve when it comes to exploiting new hardware especially something as exotic as these two units. The guy even mentions that ideally he would have like not to have to use 3 cores but you have tradeoffs that you have to make to get performance at a reasonable price.
 
Top Bottom